Hartford Foundation for Public Giving hosts conversations on using research in policy
Donaghue is in its fourth year of supporting research done with existing data through its Another Look: Better Health for Elders in Care Facilities program.
Not all questions can be answered with existing data, but many important issues can be explored in this more cost-effective way. By the end of 2016, Donaghue will have awarded $3 million for 26 projects that test better healthcare for elders. We estimate that we’d only have been able to fund half of those projects if they required new data to be collected for the analysis.
So we were interested to participate this summer with other Connecticut non-profit organizations in two meetings hosted by the Community Indicators Project of the Hartford Foundation for Public Giving about opportunities to use existing data sources and applied research to improve policy making. The Community Indicators Project works to share social and economic indicator data about life in the region, provide more targeted analysis of specific topics related to their grantmaking, and develop a supportive ecosystem for the use of data by many organizations to improve the work of the non-profit sector and municipal and state government. As one way of achieving these goals, they hosted two meetings that described the Collaborative RI and What Works CT. Each of these presentations highlights the challenges in using these data sources and the opportunities they pose for improving policy and practice.
The mission of the College and University Research Collaborative from Rhode Island, or Collaborative RI as it is known, is to increase the use of non-partisan academic research in policy making and thereby provide an evidence-based foundation for government decisions in Rhode Island. At the meeting hosted by the Community Indicator Project, Amber Caulkins, Program Director of the Collaborative RI, described its services and structure.
The Collaborative RI promotes research into action through a network of over 40 researchers at Rhode Island’s 11 public and private colleges and universities. Its leadership council includes the presidents of the schools. Policy leaders from the governor’s office and the legislature are responsible for identifying research topics. Researchers, both tenured and non-tenured faculty, receive stipends ranging from $1,000 to $10,000 and up to $25,000 for larger projects. Financial support for these stipends is shared by the state government and the participating colleges and universities, who each contribute about $7,500 per campus. In-kind support is received from the Association of Independent Colleges and Universities of Rhode Island. In addition, the Rhode Island Foundation is paying for the staff to organize the effort and its evaluation. About 20 research briefs were developed through Collaborative RI in the past year.
Caulkins said that representatives from other states, including Massachusetts, Pennsylvania, Indiana and Washington, have contacted Collaborative RI to seek information about replicating this structure in these states. More information about the work of the Collaborative RI can be found at http://collaborativeri.org.
What Works CT
The State of Connecticut’s Office of Policy and Management is starting a statewide integrated data system that will link data across all agencies in the executive branch of Connecticut’s government for use in policy analysis, program evaluation and research.
The overall goal of What Works CT is to access, link, analyze and share data maintained by executive agencies and to respond to queries from state agencies or private organizations and individuals. The Secretary of the Office of Policy and Management, who is responsible for implementing WWC, will give priority to queries that measure outcomes for state-funded programs or that may assist in the development of policies to promote the more effective and efficient use of state resources. Jamie L. Mills, Senior Advisor for Policy Analysis, and Tyler Kleykamp, Chief Data Officer from the Office of Policy and Management, described the development and potential use of the integrated data system at the meeting hosted by the Hartford Foundation’s Community Indicators Project.
Although the integrated data system will be operated by OPM, it will be informed by a Governance Committee comprised of the Commissioners of the Departments of Correction, Labor, Education and Rehabilitation Services and the Secretary of OPM. A significant role for the Governance Committee will be to prioritize the policy challenges facing the State that could benefit from integrated data by gaining a deeper understanding of service use, whether and how to increase coordination across systems, or that could enable a focus on early intervention and prevention. The Governance Committee is also responsible for establishing an advisory board of applied researchers to collaborate with state policy makers in using the state’s data to inform decisions about the improving ways to structure and deliver services.Mills, the director of the project, explained that integrated data systems usually contain person-level or encounter-level information about services provided by the state that are then linked across multiple, independent agency data systems. Mills shared one recent example of the successful use of an independent data system from Indiana, where the State wanted to better understand and address its high rate of infant mortality. A study linked data across five executive branch agencies and fifty data sets. The data revealed that the number of prenatal doctor visits was the single most significant factor in both infant mortality and low birth weight in Indiana, and this finding focused Indiana’s efforts on ensuring transportation to doctor’s visits, a problem that it had not previously understood.
What Works CT will review and approve requests for integrated data sets from agencies, non-governmental organizations or individuals. If approved, an agreement will be prepared and signed by each agency whose data is involved, as well the data user, to establish approved use of the data, data security, data confidentiality, and the named individuals granted access to the linked and de-identified data. Before the data is transferred to the requester, the data will be reviewed for quality and to determine that the data has been sufficiently de-identified.
Although challenging, the technical requirements for this data integration are not as complex as many may think. Kleykamp explained that recent changes in the State’s data systems have reduced the number of independent databases, and the majority of the information from executive branch agencies resides on three main database platforms. And although the first few data matches are expected to take several weeks to complete, de-identifying and matching algorithms are expected to be more efficient as programmers learn from previous requests.
For more information about What Works CT, contact Tyler Kleykamp at firstname.lastname@example.org.